Fetal ultrasound image brain segmentation method based on prior interaction reinforcement learning

Abstract

The invention discloses a fetal ultrasound image brain segmentation method based on prior interaction reinforcement learning, and belongs to the field of medical image segmentation and deep learning. According to the method, a fetal ultrasound image brain is segmented and converted into an environment state conversion module, an environment reward conversion module and a prompt graph updating module, and the environment reward conversion module calculates an environment reward value through label data and prediction data; the prompt graph updating module uses the prompt graph updating network to provide a new prompt graph for the current result to refine the next segmentation result; and the environment state conversion module predicts the action probability and evaluates the current state combination, and continuously iteratively updates the current state value until the segmentation result is satisfied. According to the method, the end-to-end segmentation of the fetal brain structure can be realized, the accuracy rate is relatively high, a favorable data basis and performance guarantee are provided for intelligent analysis of fetal corpus callosum development state recognition, and the method has far-reaching significance for related research of intelligent medical treatment in the future.

Description

technical field [0001] The invention relates to the field of medical image segmentation and deep learning, in particular to a method for brain segmentation of fetal ultrasound images based on prior interaction reinforcement learning. Background technique [0002] Medical image segmentation is a key technology in image analysis and processing. Separating relevant tissues of interest according to the similarity and specificity of the image area is of great significance to the clinical diagnosis and treatment process and is the main premise of all follow-up work. , the quality of the segmentation effect will directly affect the smooth progress of the information processing work. [0003] The corpus callosum is located at the floor of the interhemispheric fissure and is the largest commissural fiber in the cerebral hemisphere. Dysplasia of the corpus callosum (Agenesis of Corpus Callosum, ACC) is a congenital abnormality in fetal central nervous system malformation, which refer...

AI Technical Summary

Problems solved by technology

[0004] The automatic segmentation of brain ultrasound images is a typical task in medicine, but there is no related work on the segmentation of prenatal fetal brain ultrasound images. In order to facilitate subsequent clinical trials, this project proposes a fetal ultrasound based on prior interactive reinforcement learning for the first time. Image brain segmentation method to automatically extract fetal brain regions from ultrasound images

Images